Trazado de un pantalón.

The advantage of individual interviews is that they can provide “an in-depth understanding of farming activities, contribution to livelihoods, their constraints and aspirations” (Martin et al 2002: 7). Semi-structured interviewing was my main method because it offers flexibility both in the sense that I could ask follow-up questions, and that the interviewees could focus on the issues that are of importance to them (Bryman 2012: 487). Based on my readings and the information that I had gathered from the initial farm visits and conversations with other actors, I prepared an

interview guide (Bryman 2012; Leech 2002) with a number of topics and open-ended

questions (see Appendix 1). In most of my interviews, I aimed at creating an informal atmosphere and the feeling of having a conversation, and let the interviewees talk on the specific aspects they chose to address (as suggested by Bernard 2006). The interview guide served as a support to get back to my topics of interest if the conversation deviated too much, and to make sure that I covered everything. I asked for the interviewees’ permission to take notes and voice record the interview (a few persons declined recording), and let them select the venue. Most of the interviews with the farmers were carried out in their plot or in the collective areas or the gardens, and were not always done in privacy. Almost all of the interviews with municipal authorities (see interview guides in Appendix 2) and researchers were done in their respective offices, most of the time in privacy.

As my interviews with the urban farmers were carried out at the gardens, it became natural to make informal observations of the gardens and the farmers’ work, as well as the interaction among the farmers. I also paid attention to the interviewees’ behaviour and reactions during the interviews. My observations were kept in my field

diary, in which I also jotted down thoughts and feelings during the entire fieldwork

(Bernard 2006). I found the diary to be a very helpful tool. During the field period, it was an outlet for all the new impressions and helped me to stay focused when things got stressful or tough, and during the writing of the thesis it was an invaluable resource to remember my research process.

I also did some additional field visits to both broaden and deepen my knowledge of the case. I visited several other SMASAN initiatives in order to get a first-hand, holistic understanding of the system and its interactions. I participated in a COMUSAN meeting to learn about the social control system. I also visited two other urban gardens: the FPM-supported CEVAE ‘Morro das Pedras’, and another SMASAN-supported community garden called ‘Diamante Azul’. The purpose of these visits was to check biases and expand my empirical knowledge of urban gardens in BH, especially since I came to understand that Vila Pinho and Jardim Produtivo were two quite unique success cases and the most frequently visited. Diamante Azul is smaller and less developed, organised, and established, and I could based on the my visit and conversation with a SMASAN agronomist confirm the claims that Vila Pinho and Jardim Produtivo are indeed exceptional UA examples in Belo Horizonte.

Data analysis

During and after the fieldwork, notes and voice recordings were transcribed (Bernard 2006). My research process was iterative, with repeated data collection, analysis, coding, and specifying of research questions (Bryman 2012: 387). I have used a grounded theory coding approach, which implies “giving labels (names) to component parts that seem to be of potential theoretical significance and/or that appear to be particularly salient within the social worlds of those being studied” (Bryman 2012: 568). My labelling of data has been induced by my research questions (the topics of interest), theory (previous research), and data from my fieldwork. Codes in grounded theory are “in a constant state of potential revision and fluidity” (Bryman 2012: 568). I line with social constructivism, I have inevitably affected the coding process by determining what to include and what to omit, and how labels have been grouped, but as systems thinking (see chapter 3.2) suggests, the researcher has to set boundaries for the purpose of analysis even though everything is interrelated. The final stage of analysis entailed to interpret the codes/themes, find interrelations between them, and reflect upon their meanings and implications.

4.3 Reliability, validity, and generalizability

Since qualitative research involves people and social situations, alternative concepts have been developed to evaluate the research. I have chosen to use the four criteria suggested by Yardley (2000). First of all, Yardley emphasises sensitivity to context, both in terms of the social setting and the underlying theoretical and ethical considerations. This has been a main concern in my thesis, as I describe in the sections on social constructivism (4.1) and power relations (4.4). Second, the researcher needs to be dedicated, skilful and conduct thorough research to obtain

commitment and rigour. While I had robust background knowledge of methods and

the topic before entering the field, my skills as a researcher have evolved and improved during the process. By being transparent and keep records of all phases of the research process, I have tried to acquire transparency and coherence. Yardley also suggests that the study should be of impact and importance in practice and in theory. I have chosen a topic and an approach aiming to study and suggest improvements for a real situation, and I also believe that the multifunctionality approach can be of use in a broader setting than this specific case, as I will discuss in chapter 6.2. Yin (2009) stresses that the aim of a qualitative case study is not to generalize the findings to a population, but to theory. I thus do not say that my findings are applicable to any other setting, but that they offer an empirical example adding to the body of literature on the topic.

4.4 Reflections on my role as a researcher

Culture and language

The fieldwork period in Belo Horizonte was my first time in the city. However, I had recently lived in Brazil’s capital during five months, as well as in other parts of South America prior to that, and therefore had a sound cultural understanding of the region. Culturally, Belo Horizonte has both commonalities and differences with other parts of Brazil, but what mattered more to my study were the differences between areas within the city. In the following section (power relations, access, and trust), I will discuss how I approached different people and social settings.

Since few people speak English (even among those with high educational level) it was a necessity to speak Portuguese – all my interviews were carried out in Portuguese. My language skills were fairly good, and Brazilian people are in general encouraging. Even so, language was a challenge, because of local accents and slang that I was not used to. I considered contracting a translator, but decided not to, as I felt it would be a greater bias to my study. Instead, I asked my interviewees to clarify whenever needed, and voice recorded everything so that I could double check afterwards whether I had understood things correctly. It was a challenge to ask questions in the right way (without adding value etc.) due to my somewhat limited vocabulary. Moreover, it did sometimes obstruct the flow and asking follow-up questions, and I might at times have received incorrect or insufficient information due to my non-native culture- and language proficiency (Bernard 2006: 360-364). Yet, Bernard (2006) also stresses that some degree of naivety can be positive. I felt that being genuinely curious and interested in the interviewees encouraged them to explain things more thoroughly. Maybe, I was also more open-minded and made fewer presumptions about things compared to someone native to the context. After all, any background biases a study, but in different ways.

Power relations, access, and trust

Throughout the study, I tried to be myself, but also adapt to the different contexts (Bernard 2006: 358). Sultana (2007) emphasises that “it is important to pay greater attention to issues of reflexivity, positionality and power relations in the field in order to undertake ethical and participatory research” (Sultana 2007: 374). My own cultural background and ethnicity, as European, white, young, female, and academic, are likely to have affected the study (Sultana 2007). She also notes that going abroad sometimes means ascending to another social class; while I would consider myself middleclass in Sweden, I would probably be positioned as upper-middleclass in Brazil, which is important to be aware of. Sultana also underlines the fact that power relations “influences methods, interpretations, and knowledge production”, and it has been important to me to reflect upon the information I gathered: what information and opinions did the interviewees share with me, and why?

It was easier than I had expected to gain access to the municipal authorities, and I was amazed by people’s willingness to take time to talk to me and show me around. Depending on the person’s hierarchy within the department/institution, I could see differences in access and the degree of planning and formality an interview required, but I was always met with a welcoming attitude. I am not sure what enabled my access, but I do have some reflections. First of all, I believe that my key informant (an UFMG researcher) really was key in gaining access to the municipal authorities. Thanks to his network, I immediately gained trust from them, and they in turn “validated me” to others. Secondly, my background and ethnicity might have helped, although my key informant claimed that the departments are rather open and that a Brazilian student would gain access with the same ease. While that might be true, it might not be the case for a Brazilian person belonging to a different societal class. In terms of gaining trust among the farmers, my background could rather be a disadvantage precisely because of underlying inequalities and power relations. First of all, I made sure to explain who I am, why I was there and what I wanted to study. While the civil servant from SMASAN who introduced me to the urban farmers was a necessary gate-opener and gave me legitimacy, I had to clarify that I was an independent student and not affiliated with the municipality (Bryman 2012: 439). I explained that I also conducted interviews with municipal authorities, that I was there not to assess them but to learn, and that all interviews were equally important. I also underlined the voluntary nature of participation and that they would remain anonymous in my study. I occasionally felt uneasy with not being able to give a lot back. Some of the farmers meant that researchers just extract information and leave, and that they need more resources, and one farmer declined participation because she was tired of telling the story of the garden once again, which probably is due to the fact that these two gardens are the most visited ones, as I came to understand during my fieldwork. However despite the short time in the field, I felt that I managed to gain people’s acceptance and trust. While I knew that my background posed challenges, I believe that finding a common language, focusing on similarities (in my case for instance, being from the countryside), and sharing personal stories and thoughts, was key to building trust. I also think that more “obvious” things such as being respectful, open-minded, flexible, and humble go a long way. I got several indications of acceptance; one woman that I interviewed told me that she normally does not talk to people, some of the other farmers invited me to share their lunch and visit their homes, and another told me that he appreciated my respect towards them and the mere fact that I was talking directly to those who do the farming and not only to those higher up in the hierarchy. Finally, I made sure to inform the farmers of the length of my fieldwork, and made a final visit to say goodbye before leaving the field (Bernard 2006: 383).

5. Results

This chapter presents data gathered from my fieldwork in Belo Horizonte. The following research questions have been studied:

1) How are the urban gardens set up and managed? What are the main challenges? 2) What are the benefits of the urban gardens, according to the urban farmers? 3) How is UA framed in local politics in Belo Horizonte?

• Which is the political home of UA?

• How do different municipal government bodies approach UA?

• How is the UA work of different municipal government bodies integrated?

5.1 Farming system of the studied gardens

This subchapter presents the findings for research question 1: How are the urban

gardens set up and managed? What are the main challenges?

Vila Pinho (picture 2) was the first garden that SMASAN initiated. It is also the

largest, with 19 farmers sharing 10 000 m2. It was created in 1998 to improve people’s food and nutrition security and the safety in the area. Jardim Produtivo (picture 1) was created in 2008 as a pilot garden within the CFF project. Today there are ten farmers at Jardim Produtivo. Figure 3 uses the adapted version of the Hawkesbury’s Peanut Model (introduced in chapter one), to present a combined image of the farming systems at Vila Pinho and Jardim Produtivo15.

15 _{I chose to present a combined image for reasons of clarity and because the two gardens’ farming}

systems and the challenges they face are very similar.